Hardware-aware comparative study of lightweight convolutional neural networks for Raspberry Pi-based autonomous driving.
Saved in:
| Title: | Hardware-aware comparative study of lightweight convolutional neural networks for Raspberry Pi-based autonomous driving. |
|---|---|
| Authors: | Kim, Hyung In1 khi103@yonsei.ac.kr, Park, Youngmin2 ympillow@sookmyung.ac.kr |
| Source: | International Journal of Electrical & Computer Engineering (2088-8708). Jun2026, Vol. 16 Issue 3, p1493-1507. 15p. |
| Subjects: | Raspberry Pi, Convolutional neural networks, Autonomous vehicles, Computer storage capacity, Regression analysis, Model validation, Computer peripherals |
| Abstract: | Deploying deep learning models for autonomous driving on resourceconstrained edge devices, such as the Raspberry Pi, presents significant challenges due to strict limitations on inference latency and memory capacity. To address these constraints, this study conducts a comprehensive comparative evaluation of lightweight convolutional neural networks (CNNs) optimized for dual-output regression of steering angle and driving speed. We benchmark a task-specific end-to-end baseline (NVIDIA CNN) against representative classification-oriented architectures--including MobileNet, ShuffleNet, EfficientNet, GhostNet, and SqueezeNet--all reformulated for this regression task. Experiments were conducted on a physical Raspberry Pi-based autonomous RC car platform to assess prediction accuracy, inference speed, and real-world closed-loop driving stability using quantitative metrics such as the normalized jerk ratio. Experimental results demonstrate a clear trade-off: while Ghost-NetV1 0.5× achieved the highest regression accuracy with a Total R2 score of 95.8% and MobileNetV1 recorded a competitive MAE of 1.95, they failed to provide stable control due to severe high-frequency steering jitter. Conversely, the NVIDIA CNN proved to be the most practical solution for general edge deployment, achieving the lowest inference latency of 61.1 ms (16.4 FPS) and a minimal memory footprint of 2.78 MB, ensuring stable autonomous navigation (1.50× jerk ratio). Furthermore, ShuffleNetV2 0.5× emerged as the superior architecture for trajectory precision, recording the lowest weighted MAE of 1.60. These findings underscore that theoretical accuracy does not guarantee real-world drivability on embedded systems, providing practical guidelines for hardware-aware model selection in edge-based autonomous driving. [ABSTRACT FROM AUTHOR] |
| Copyright of International Journal of Electrical & Computer Engineering (2088-8708) is the property of Institute of Advanced Engineering & Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Engineering Source |
| FullText | Links: – Type: pdflink Text: Availability: 0 |
|---|---|
| Header | DbId: egs DbLabel: Engineering Source An: 194285636 AccessLevel: 6 PubType: Academic Journal PubTypeId: academicJournal PreciseRelevancyScore: 0 |
| IllustrationInfo | |
| Items | – Name: Title Label: Title Group: Ti Data: Hardware-aware comparative study of lightweight convolutional neural networks for Raspberry Pi-based autonomous driving. – Name: Author Label: Authors Group: Au Data: <searchLink fieldCode="AR" term="%22Kim%2C+Hyung+In%22">Kim, Hyung In</searchLink><relatesTo>1</relatesTo><i> khi103@yonsei.ac.kr</i><br /><searchLink fieldCode="AR" term="%22Park%2C+Youngmin%22">Park, Youngmin</searchLink><relatesTo>2</relatesTo><i> ympillow@sookmyung.ac.kr</i> – Name: TitleSource Label: Source Group: Src Data: <searchLink fieldCode="JN" term="%22International+Journal+of+Electrical+%26+Computer+Engineering+%282088-8708%29%22">International Journal of Electrical & Computer Engineering (2088-8708)</searchLink>. Jun2026, Vol. 16 Issue 3, p1493-1507. 15p. – Name: Subject Label: Subjects Group: Su Data: <searchLink fieldCode="DE" term="%22Raspberry+Pi%22">Raspberry Pi</searchLink><br /><searchLink fieldCode="DE" term="%22Convolutional+neural+networks%22">Convolutional neural networks</searchLink><br /><searchLink fieldCode="DE" term="%22Autonomous+vehicles%22">Autonomous vehicles</searchLink><br /><searchLink fieldCode="DE" term="%22Computer+storage+capacity%22">Computer storage capacity</searchLink><br /><searchLink fieldCode="DE" term="%22Regression+analysis%22">Regression analysis</searchLink><br /><searchLink fieldCode="DE" term="%22Model+validation%22">Model validation</searchLink><br /><searchLink fieldCode="DE" term="%22Computer+peripherals%22">Computer peripherals</searchLink> – Name: Abstract Label: Abstract Group: Ab Data: Deploying deep learning models for autonomous driving on resourceconstrained edge devices, such as the Raspberry Pi, presents significant challenges due to strict limitations on inference latency and memory capacity. To address these constraints, this study conducts a comprehensive comparative evaluation of lightweight convolutional neural networks (CNNs) optimized for dual-output regression of steering angle and driving speed. We benchmark a task-specific end-to-end baseline (NVIDIA CNN) against representative classification-oriented architectures--including MobileNet, ShuffleNet, EfficientNet, GhostNet, and SqueezeNet--all reformulated for this regression task. Experiments were conducted on a physical Raspberry Pi-based autonomous RC car platform to assess prediction accuracy, inference speed, and real-world closed-loop driving stability using quantitative metrics such as the normalized jerk ratio. Experimental results demonstrate a clear trade-off: while Ghost-NetV1 0.5× achieved the highest regression accuracy with a Total R2 score of 95.8% and MobileNetV1 recorded a competitive MAE of 1.95, they failed to provide stable control due to severe high-frequency steering jitter. Conversely, the NVIDIA CNN proved to be the most practical solution for general edge deployment, achieving the lowest inference latency of 61.1 ms (16.4 FPS) and a minimal memory footprint of 2.78 MB, ensuring stable autonomous navigation (1.50× jerk ratio). Furthermore, ShuffleNetV2 0.5× emerged as the superior architecture for trajectory precision, recording the lowest weighted MAE of 1.60. These findings underscore that theoretical accuracy does not guarantee real-world drivability on embedded systems, providing practical guidelines for hardware-aware model selection in edge-based autonomous driving. [ABSTRACT FROM AUTHOR] – Name: AbstractSuppliedCopyright Label: Group: Ab Data: <i>Copyright of International Journal of Electrical & Computer Engineering (2088-8708) is the property of Institute of Advanced Engineering & Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract.</i> (Copyright applies to all Abstracts.) |
| PLink | https://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=egs&AN=194285636 |
| RecordInfo | BibRecord: BibEntity: Identifiers: – Type: doi Value: 10.11591/ijece.v16i3.pp1493-1507 Languages: – Code: eng Text: English PhysicalDescription: Pagination: PageCount: 15 StartPage: 1493 Subjects: – SubjectFull: Raspberry Pi Type: general – SubjectFull: Convolutional neural networks Type: general – SubjectFull: Autonomous vehicles Type: general – SubjectFull: Computer storage capacity Type: general – SubjectFull: Regression analysis Type: general – SubjectFull: Model validation Type: general – SubjectFull: Computer peripherals Type: general Titles: – TitleFull: Hardware-aware comparative study of lightweight convolutional neural networks for Raspberry Pi-based autonomous driving. Type: main BibRelationships: HasContributorRelationships: – PersonEntity: Name: NameFull: Kim, Hyung In – PersonEntity: Name: NameFull: Park, Youngmin IsPartOfRelationships: – BibEntity: Dates: – D: 01 M: 06 Text: Jun2026 Type: published Y: 2026 Identifiers: – Type: issn-print Value: 20888708 Numbering: – Type: volume Value: 16 – Type: issue Value: 3 Titles: – TitleFull: International Journal of Electrical & Computer Engineering (2088-8708) Type: main |
| ResultId | 1 |
